Présentation des prompts et skills

Prompts IA sélectionnés pour des usages professionnels, créatifs et techniques.

Works well with Claude, GPT-4o/5, or Gemini — models strong at structured financial reasoning and multi-step tradeoff analysis. A longer-context model helps if your budget has many line items.Finance et NégociationFinance just asked you to find $80,000 in savings from your department's budget for next quarter. You have a spreadsheet with 22 line items — software licenses, contractor spend, a training budget, a deferred infrastructure upgrade — and you need a defensible recommendation by end of day, not just a list of the biggest numbers.

L'Optimiseur de Budget : Transformez un Budget Détaillé en Arbitrages Court Terme vs Long Terme

Un prompt qui évalue chaque poste budgétaire sur deux axes — économies immédiates et coût caché à long terme — et signale les coupes qui semblent bonnes maintenant mais coûtent plus cher plus tard, avec le point d'équilibre indiqué.

9/17/2026

prompt-engineeringfinance

Works well with Claude, GPT-4o/5, or Gemini — any model strong at maintaining consistent tone across variants. A longer-context model helps if the original message is lengthy.Rédaction et CommunicationYou need to tell a long-standing client their project is delayed by two weeks. The client has pushed back hard on deadlines before, and you have five minutes before a scheduled call to decide how bluntly to deliver the news.

Le Transformateur de Ton : Transformez Un Message en Quatre Versions Prêtes à Envoyer

Un prompt qui réécrit tout message à enjeux élevés en quatre tons distincts — Direct, Diplomatique, Enthousiaste, Formel — pour que choisir le bon registre prenne des secondes plutôt que 15 minutes d'hésitation.

9/17/2026

prompt-engineeringproductivity

Claude Opus 4.8 or GPT-5.4 (both hold character voice consistently across a full dialogue); avoid smaller/faster models for this one, they tend to flatten the two characters into the same voicecreativeA developer relations writer at an AI company has been asked to add a plain-English explanation of why language models hallucinate to the new-hire onboarding doc, after the third all-hands where a non-technical new employee asked why the AI just made something up. The existing explanation is three paragraphs of next-token-prediction jargon that nobody reads past the first sentence, and leadership wants something people will actually remember when a customer asks them the same question.

L'Explicateur Improbable : Transformer un Sujet Complexe en un Dialogue Entre Deux Personnages Qui Ne Se Rencontreraient Jamais

Un prompt qui oblige une explication à mériter sa place grâce à la voix des personnages et à de vraies analogies, plutôt qu'un paragraphe chargé de jargon que personne ne relit jamais.

9/15/2026

communicationonboarding

Claude Opus 4.8 (best for long-context synthesis across many sources); GPT-5.4 and Gemini 3 Pro also handle this well up to 5-6 sourceslearningA product manager at a mid-size software company has been asked by leadership to recommend whether to expand AI coding assistant licenses from a 20-person pilot team to all 400 engineers. They have pulled five research reports and internal surveys that reach opposite conclusions about whether the tools actually help experienced developers, and need to brief the VP of Engineering by Friday without just picking whichever study confirms what leadership wants to hear.

Le Cartographe de Contradictions de Sources : Transformez des Recherches Contradictoires en une Image Claire de ce que l'on Sait Réellement

Un prompt structuré qui refuse de faire la moyenne des désaccords entre sources — il indique précisément où elles s'accordent, où elles se contredisent, et pourquoi, afin que vous puissiez prendre une décision défendable.

9/15/2026

decision-makinglearning

Works well with any strong reasoning model — Claude Opus 5, GPT-5.4, or Gemini 3 Pro. Since this task involves inferring likely causes from incomplete or vague data rather than pure extraction, models with stronger reasoning tend to produce better-calibrated confidence levels on their inferences than smaller, faster models.Startups et EntreprisesA SaaS founder pulls 40 cancellation survey responses from the last quarter. Most cite "too expensive" or "didn't need it," but the churn rate keeps climbing even after a recent price cut — a sign the stated reasons aren't the real ones, and there's no time to manually cross-reference every response against usage data before the next board meeting.

L'Investigateur des Causes de Churn : Transformer les Feedbacks d'Annulation en une Liste de Correctifs Priorisés pour la Rétention

Un Prompt structuré qui distingue le churn remédiable du churn irrémédiable, creuse au-delà des raisons d'annulation vagues pour en identifier les vraies causes, et débouche sur des correctifs de rétention spécifiques et testables plutôt que sur une liste de plaintes.

9/14/2026

prompt-engineeringsaas

Works best with reasoning-focused models (Claude Opus 5, GPT-5.4, Gemini 3 Pro). For highest reliability, run the audit with a DIFFERENT model than the one that generated the original content — cross-model review catches fabrications a single model is blind to, since a model rarely flags the kind of confident-sounding claim it would have generated itself.Intelligence ArtificielleA marketing manager used ChatGPT to draft a competitive analysis report due to the VP of Sales in 45 minutes. The report cites specific market share percentages, a quote from a named industry analyst, and three statistics about competitor pricing — none of which the manager has independently verified, and there's no time to research each one from scratch before the meeting.

L'auditeur de fiabilité des sorties IA : transformez n'importe quel rapport généré par l'IA en brouillon vérifié et validé

Un Prompt d'audit structuré qui classe chaque affirmation d'un document rédigé par l'IA comme vérifiable, invérifiable ou suspecte — plutôt que de simplement demander à un modèle de « revérifier » son propre travail.

9/14/2026

prompt-engineeringAI reliability

Claude Opus 4.6 (works well with GPT-5.4 and Gemini 3 Pro — any model with strong instruction-following for structured comparison tasks performs well)CarrièreYou found a job posting for a role that's a stretch — one level above your current title — and you have 45 minutes before the application closes tonight. You need to know fast whether your resume as-is will get filtered out, and what to fix first if it will.

L'Analyseur d'Écarts de CV : Transformez une Offre d'Emploi en Liste Priorisée de Corrections

Comparez votre CV à une offre d'emploi spécifique ligne par ligne et obtenez des écarts classés et des réécritures XYZ avant de soumettre.

9/12/2026

prompt-engineeringCareer

Claude Opus 4.6 (works well with GPT-5.4 and Gemini 3 Pro — any model with strong reasoning over role/priority inference performs well)productivityYou have a 30-minute call in two hours with a VP of Engineering at a potential client and two of their staff engineers, and all you have is their names, titles, and a two-line agenda emailed by an assistant. You've never spoken with any of them before.

Le sprint de préparation de réunion : transformez des noms de participants et un ordre du jour en briefing de cinq minutes

Transformez quelques noms, titres et un ordre du jour vague en un briefing de cinq minutes qui révèle la véritable objection avant le début de la réunion.

9/12/2026

prompt-engineeringproductivity

Claude Opus 5 (strong at holding multiple data points in context while distinguishing statistically significant trends from noise); GPT-5.4 and Gemini 3 Pro also handle this well, though Claude is more consistent about flagging small-base percentage distortions unprompted.Analyse de DonnéesYou're a data analyst at a mid-size e-commerce company. Q3 sales data just landed in a spreadsheet with dozens of columns across 5 regions, and your VP of Sales wants the 3 things that matter by end of day — not a dashboard, not a slide deck, just the insights.

Le Convertisseur de Données Brutes en Note Exécutive : Transformez des Chiffres Désordonnés en Analyse en 4 Étapes Prête à l'Action

Un prompt structuré qui force le modèle à passer par les tendances, anomalies, implications commerciales et recommandations classées — au lieu de simplement décrire ce que montrent les données.

9/11/2026

data analysisanalytics

Claude Opus 5 (strong at generating genuinely distinct psychological angles rather than superficial rewordings of the same hook); GPT-5.4 and Gemini 3 Pro also work well, though Claude tends to produce more varied trigger coverage without a follow-up prompt asking it to differentiate further.marketingYou're a content marketer at a B2B SaaS company with 45 minutes before Monday's content calendar meeting. You need headline options for next week's blog post launch, but the three drafts you've already written all sound like slight rewordings of the same idea.

Le Générateur d'Accroches de Titres : Transformez Un Sujet en 20 Déclencheurs Psychologiques Étiquetés

Un prompt structuré qui produit 20 variantes de titres réparties sur six déclencheurs psychologiques — écart de curiosité, spécificité, aversion à la perte, cadrage contraire, urgence et identité — chacune étiquetée avec la raison de son efficacité pour votre audience.

9/11/2026

marketingcopywriting

Claude Opus 5 (best for reasoning through non-obvious edge cases and generating well-structured test code); also works well with GPT-5.4 or Gemini 3 Pro for most languages, though Claude tends to catch more subtle edge cases in code-review-style tasks.Outils de DéveloppementYou're a backend developer at a mid-size e-commerce company. You just finished writing a discount-calculation function for the checkout flow, code review is in 20 minutes, and you haven't written a single test yet.

L'Auditeur de Couverture de Tests : Transformez Toute Fonction en Suite pytest Complète Avant la Revue de Code

Un prompt structuré qui transforme une fonction et son contexte d'appel en tests de chemin nominal, cas limites, scénarios d'erreur et tests d'intégration.

9/10/2026

testingdeveloper tools

Works well with any strong reasoning model — Claude Opus 4.8, GPT-5.4, or Gemini 3 Pro. This task benefits from a model that generates multiple plausible interpretations rather than converging early on one guess — if a response feels too confident after only one pass, explicitly ask it to generate at least two alternative readings before continuing.CarrièreYou just left a performance review or 1:1 with feedback like 'be more strategic' or 'show more executive presence' — phrases that sound important but give you no idea what to actually change before the next review cycle.

Le Décodeur de Feedback du Manager : Transformez un Retour Vague en Plan d'Action Concret

Un prompt qui génère plusieurs interprétations concrètes d'un retour ambigu comme « sois plus stratégique », identifie la plus probable, et construit un plan mesurable sur 30-60-90 jours.

8/20/2026

Careercommunication